Aftertouch v/s mod wheel control for velocity

[Neole]

Isnt controlling velocity via mod wheel something that could be done with aftertouch as an alternative? What would be the difference, or advantage of using a mod wheel over aftertouch?

[dewdman42]

I used to have this DW-8000 that had great aftertouch. It was one of the first keyboards to feature aftertouch and I never played another one again that was as good as that Korg DW-8000 for giving me a sense of expression. One patch in particular I remember..this brass patch...as I pressed down harder on the notes they got brassier and louder and a little vibrato kicked in. It was very expressive and for me it was easier to control in a musically expressive way than I find using the mod wheel to be.

However, many keyboards don\'t have aftertouch. Isn\'t it just another CC#? You should be able to find a way to remap it to #1 if you wanted to try it.

[Markleford]

Originally posted by dewdman42:
Isn\'t it just another CC#? You should be able to find a way to remap it to #1 if you wanted to try it. <font size=\"2\" face=\"Verdana, Arial\">Aftertouch is not a CC, but an entirely different MIDI message entirely. Which itself is different from \"Channel Pressure\", which is what is implemented most often in keyboards (I think).

Aftertouch is transmitted on a per-key basis, so that really isn\'t applicable to CC situations (unless you just want to average the messages over a timeslice, which essentially produces channel pressure-like data!).

However, both these messages could be translated to CC\'s, given a flexible system (such as MFX capability)

- m

[Neole]

Originally posted by Christian Obermaier:
can\'t start a note with pressure half way or full up, and you can\'t end a note at anything else but pressure zero. This will be a very limiting factor.

Christian <font size=\"2\" face=\"Verdana, Arial\">You just press the key down to the desired level very quickly and you would be starting at that perssure, not \'fading\' into that pressure. Same with ending the note, just release the key very quickly and you wont be fading to zero pressure. Or would there be some audible difference between that and doing it with the mod wheel?

[Tom Hopkins]

There are numerous reasons (many of them given above) why aftertouch is not a good choice for this function. I’ll add one more: It is one of the most inconsistently implemented of the available MIDI controllers. By that I mean there is enormous variation between hardware controllers in the way aftertouch responds. I have a keyboard controller, for instance, that gives almost no control in the 0-64 range of travel. No matter how carefully you try to add (or reduce) pressure the data will jump in large increments (10 or more). Dynamic control of volume/timbre changes requires considerably higher resolution to function well. And yes, there is unavoidable interaction between velocity and aftertouch which would be extremely detrimental to the design goals of GPO. Aftertouch is just the wrong controller for the job.

Tom

P.S. Neole’s original question says, “Isn’t controlling velocity via mod wheel something . . .” Mod Wheel doesn’t control velocity, it controls continuous changes in individual instrument volume/timbre. It roughly substitutes for cc#11 in the MIDI spec but goes beyond that by adding other dynamic changes too (timbre). In other libraries velocity is often used for volume changes but only at the initial note “on.” Velocity is not capable of controlling crescendos and diminuendos within notes (or groups of notes) as the mod wheel does in GPO. GPO uses velocity in standard fashion only for instruments that are percussive in nature. Expressive instruments use velocity to control attack strength.

[nexus]

Christian and Tom know what they are talking about. Aftertouch is a misnamed function, it used to be referred to as \'poly aftertouch\'. The way \'channel pressure\' is implemented is wacky in many keyboards showing the lack of interest by manufacturers in developing it further.

Channel pressure would really limit you in the kinds of smooth expressive lines you could get. It might be cool in a synth like the DW-8000 (an underrated instrument IMO), but very unatural for a likelike instrumental phrase (to say nothing of the increased CPU load in your sequencer all that cc is throwing out!).

Orchestras change dynamics with a more flowing motion.

Some good advice I read many years ago was to think in terms of the instruments and sections of an orchestra as \"always getting louder or always getting softer\". The orchestra therefore, is constantly changing and moving. This is the key to realistic performances as the better GPO demos exhibit. Seldom will a composer request \'stagnant tone\' from an instrument.

This constantly changing dynamic would be hard to accomplish smoothly (or at all) with channel pressure.

[JAK]

[QUOTE=Tom Hopkins]There are numerous reasons (many of them given above) why aftertouch is not a good choice for this function. I’ll add one more: It is one of the most inconsistently implemented of the available MIDI controllers. By that I mean there is enormous variation between hardware controllers in the way aftertouch responds. I have a keyboard controller, for instance, that gives almost no control in the 0-64 range of travel. No matter how carefully you try to add (or reduce) pressure the data will jump in large increments (10 or more). Dynamic control of volume/timbre changes requires considerably higher resolution to function well. And yes, there is unavoidable interaction between velocity and aftertouch which would be extremely detrimental to the design goals of GPO. Aftertouch is just the wrong controller for the job.

I certainly will not dispute your technical explanations regarding aftertouch as not viable as a controller.

I will say however, that I compare it directly to a violinist who may use velocity ( speed of bow) and pressure ( aftertouch ) .

I feel both would work very well IF technically it could be achieved. Having read your post I can readily see the problems.

I have the aftertouch on one of my controllers and of course the modwheel for GPO.

Given the choice if viable? I would overwhelmingly choose the aftertouch because I can FEEL the crescendo .

With a modwheel crescendo you can HEAR it but not feel it in my experience.

To me it's like adding sustain to a piano passage ...later. Works but doesn't feel nice at all.

I wish we had the choice, I really do, but I understand you must make decisions that benefit most people and for the reasons you stated above.

Respectfully JAK

[Bela D Media]

With DIVA - aftertouch is one way to control the vibrato samples in sync with the straight tone samples.
We give the option of mod wheel or aftertouch.

The "hands on" approach "feels" more expressive than mod wheeling but some do not have a MIDI controller that transmits aftertouch - Radium 61 for example - so both options are available.


Aftertouch - an important yet overlooked control setting

Our 2 cents ;)

[Alan Lastufka]

The biggest problem with aftertouch is its lack of support on most keyboards... including mine... Gary considered it for GPO, but after finding out that it would work on mine I talked him out of it... J/K

However, it's lack of support by many keyboard companies is disheartening, as it would be a really cool controller.

[JAK]

The biggest problem with aftertouch is its lack of support on most keyboards... including mine... Gary considered it for GPO, but after finding out that it would work on mine I talked him out of it... J/K

However, it's lack of support by many keyboard companies is disheartening, as it would be a really cool controller.


Yes it really is disheartening. I thought about this last night and this morning. I am a string specialist having composed for real orchestra and hearing it performed and of course many midi productions.

Furthermore I'm a bit of a baroque specialist as well.

The velocity works amazingly well but my dream would be to hit a sustain at any time as well and simply press down to crescendo .......just beautiful.

When I use the mod wheel ...yes it works well...it just doesn't FEEL well....
and aside from that I would love to be able to, if there is no aftertouch available, to be able to use the exp control for a few notes and then suddenly do a faster few notes with a harder touch a/la velocity .

I started with the GPO and shortly after that ordered GOS Lite.
Suitably impressed I will now be ordering the GOS Full.

Currently I layer 4 deep in Gigastudio 160. 64 channels comprised of Garritan strings mixed with Dan Dean Solo strings.

Layer 1 GOS sustain, layer 2 is an attack or spiccato, layer three is a Dan Dean solo spiccato, and layer 4 is Dan Dean solo arco.

This will not work with GOS full . I will need Giga 3.0 for up to 8 layers.
Therefore I will sometimes use 4-6 layers.

Again , I think the results would be astounding with the velocity/aftertouch combination.

Even if we composed a short list of controllers for the forum for those interested the libraries themselves would have to offer this option.

[billp]

This is complicated by the fact that there is both key aftertouch and channel pressure in the midi spec...but assuming key aftertouch...

This looks like another MFX project for Markleford.

Sounds to me like the solution is to have aftertouch generate incremental MOD values, but never to cause the MOD level to drop below the value set by the MOD wheel itself. For example if MOD is 100, any after touch value that exceeds 100--110, for example--would transmit a MOD message of 110--100 (from the MOD wheel) + 10 (the delta from aftertouch). If the aftertouch was less than the last MOD value transmitted (100 in the example), it would have no effect.

This is complicated by the fact that key aftertouch is per key, so in a multi-voice situation, the highest aftertouch value would end up governing the entire channel/voice. But for strings and brass, this could still operate fairly consistently.

Its 5:30 am...maybe this won't work at all and I just need coffee.

[Markleford]

Okay, I made an MFX plugin that does something to this effect. Check out this other thread (http://www.northernsounds.com/forum/showthread.php?t=21971)! :)

- m